**** BEGIN LOGGING AT Sun Jul 23 02:59:57 2006 Jul 23 05:05:36 the wiki said to pull org.openembedded.dev so i did - but further down to set up local conf it seems to reference org.openembedded.oz354x - so what is correct here? and can you guys help update the wiki (im willing to edit if i know what is right) Jul 23 05:06:33 but i think the simplest thing would be for one of you gurus to update the wiki and then just let me read it - like the whole killing two birds with one stone thing Jul 23 05:07:20 anyway, since im stumped, im going to bed and i will come back and read your responses in the (eastern time) morning - thanks Jul 23 05:46:20 poushag: use 354x as stable and .dev as unstable Jul 23 06:31:27 good morning all Jul 23 06:42:51 hi koen - the wiki is messed up - says to pull .dev - then later says to use .354x Jul 23 06:43:32 if i try to fix it it may get worse Jul 23 06:44:00 but maybe you can look at the getting started page and make the appropriate edits Jul 23 06:44:41 ok im really cutting off the computer (screen) now - good night Jul 23 06:46:32 http://www.openembedded.org/wiki/GettingStarted never even mentions oz354x Jul 23 07:05:52 http://www.openembedded.org has a new frontpage story saying that people can start pushing csets again Jul 23 07:09:54 cool Jul 23 07:10:09 koen: did you see my response re minisplat video? Jul 23 07:11:33 i've got bb files for it that do not yet have video; my local oe tree has a new ffmpeg_svn.bb file and a fairly substantial new patch for their configure script Jul 23 07:13:12 i need to recruit some others to help me test on zaurus/other oe targets Jul 23 07:13:46 ah, cool Jul 23 07:14:19 this week, i got distracted implementing a new pastebin client/server system Jul 23 07:15:06 it now lets me paste the output from 'svn diff' (or --diff-cmd) to the server in one step, then i can apply the patch from the server on a repo on another machine; all over http Jul 23 07:16:22 i was thinking about rewriting bitbake's pastebin capabilities to use it, since i'm already hooked on using it for moving patches to/from my openbsd test box (behind a firewall and - when i started this - w/ sendmail) Jul 23 09:10:56 morning all Jul 23 09:11:28 hey RP Jul 23 09:11:38 how do you like the speed on the new monotone? Jul 23 09:12:09 koen: Its a definite improvement :) Jul 23 10:52:15 RP: you might consider running 'mtn refresh_inodeprints' inside the workspace Jul 23 10:53:32 koen: What makes you suggest that? Jul 23 10:53:57 past complaints from o-handians about monotone slowness Jul 23 10:57:02 koen: I was wondering if my commits had gone wrong somehow :) Jul 23 10:57:53 heh Jul 23 10:57:58 see http://www.venge.net/monotone/docs/Inodeprints.html for the docu Jul 23 10:58:11 on one machine it helps, on the other it doesn't Jul 23 10:59:40 I suspect it will help a lot on my machines :) Jul 23 11:02:10 * koen stops fiddling with apache Jul 23 11:04:55 hey pH5 Jul 23 11:05:07 hey koen Jul 23 11:21:45 ~lart exim, dns, mx records and e-mail in general Jul 23 11:21:46 * ibot_ teaches exim, dns, mx records and e-mail in general that M$ Access is a database. No, really, a database. A real live multi-user... well, ok, not multi-user, but a database. Yeah, that sounds right. Jul 23 11:22:14 @openembedded.org mail addresses are *almost* working Jul 23 11:23:11 koen: You should be able to use the rewrite section in exim's config to do that.. Jul 23 11:23:50 adding a correct mx recored would be more :) Jul 23 11:24:16 I'll leave that to mickeyl and kergoth Jul 23 11:26:03 Ah ;) Jul 23 11:50:36 morning Jul 23 11:51:07 hey chouimat Jul 23 11:51:31 afternoon all Jul 23 11:51:49 hey ade|home Jul 23 11:52:05 if the news on oe.org is correct should the title here be updated ? Jul 23 11:53:01 I guess something went wrong the last time I tried that Jul 23 11:53:11 nicely done sir Jul 23 11:54:00 hippy dmz machine at work has dropped off the network so can't download the snapshot Jul 23 11:54:06 what a shame Jul 23 11:54:19 initial pull is quite fast now :) Jul 23 11:54:44 nice to hear. Jul 23 11:55:24 i like the snapshot as i can install it on my machine at home and not use up my download limit on adsl Jul 23 11:55:46 * chouimat bangs his head on desktop screamming shit they don't know how to understand what they are reading Jul 23 11:56:07 I have replaced oe.db with oe.mtn and installed monotone version 0.28. Going to org.openembedded.dev and doing a "mtn update" gives the following error "mtn: misuse: workspace required but not found" Jul 23 11:56:10 03pH5 07org.oe.dev * r07af0005... 10/ (9 files in 5 dirs): anthy: update to 7811 Jul 23 11:56:14 Google does not know about this error. Jul 23 11:56:16 03pH5 07org.oe.dev * r50eda023... 10/ (6 files in 4 dirs): tango-icon-theme: add 0.7.2 and -extras 0.1.0 with icon-naming-utils disabled Jul 23 11:56:16 before any of you ask here what I'm talking about http://www.kdedevelopers.org/node/2187 Jul 23 11:56:22 03pH5 07org.oe.dev * r40fe085c... 10/ (5 files in 3 dirs): tomoe-0.2.1: add tomoe, a handwriting recognition engine Jul 23 11:56:27 yay, write access :) Jul 23 11:56:27 03pH5 07org.oe.dev * re789750d... 10/ (7 files in 5 dirs): hdparm: add 6.6 Jul 23 11:56:28 pH5: Nice to see the update coming. Jul 23 11:56:32 03pH5 07org.oe.dev * rafd75613... 10/ (8 files in 3 dirs): bluez: add bluez-hcidump-1.31, bluez-{libs,utils}-3.1 Jul 23 11:56:41 Laibsch: you need to do a new checkout Jul 23 11:56:52 pH5: Your work? Or stuff from Tota? Jul 23 11:56:59 koen: OK. Jul 23 11:58:24 oh good point, whats the way to ack a patch from another source in OE? Jul 23 11:58:58 or do we just put who made it in the commit log? Jul 23 11:59:46 Laibsch: not sure, it was in my tree for quite some time. can you point me to tota's patches for anthy and tomoe? Jul 23 12:00:28 ade|home: mtn commit --author Jul 23 12:00:30 pH5: all I remember from the top off my hat is bug 1094, meta bug for all stuff CJKV. Take a look at the open ones. Jul 23 12:01:05 ade|home: or put it in the patch header and commit log Jul 23 12:01:20 pH5: I think anthy has been in OE for some time now. Jul 23 12:01:29 Laibsch: ok, will do. Jul 23 12:01:50 Does your stuff include imkit? Without it is more or less useless. Jul 23 12:02:13 I think tomoe is completely new in OE (which would be great news). Jul 23 12:03:29 Laibsch: it is. I didn't touch imkit as it's opie and mickey said there were build issues. Jul 23 12:04:05 ade|home: check http://www.openembedded.org/wiki/MonotonePhraseBook :) Jul 23 12:04:13 Is it opie? I think it might be/should be possible to run it regardless of window manager. Jul 23 12:04:26 opie isn't a wm Jul 23 12:04:37 it's using that qt/e atrocity Jul 23 12:04:51 pH5: Yes, unfortunately, nobody here in OE was able to build imkit yet. Jul 23 12:05:25 tota-san from openzaurus-ja has got it compiling, though. His stuff put straight into OE does not compile. Jul 23 12:05:55 koen: Whatever the name, the point remains. Jul 23 12:08:48 hmmmm Jul 23 12:08:54 pH5: Does your stuff fix bug 1132? Jul 23 12:09:08 koen: cheers Jul 23 12:10:47 migration on my humble system went painlessly -- I agree, well done! Jul 23 12:11:09 Bloody hell.. monotone 0.27's fast. Jul 23 12:12:16 Laibsch: it does install /usr/share/anthy.dic, can you test if it fixes 1132? Jul 23 12:12:52 I was never sure what tota meant. Jul 23 12:13:23 I will give it a spin later. Jul 23 12:13:29 Laibsch: thanks Jul 23 12:13:34 I think I have some general idea what it might be. Jul 23 12:13:46 pH5: Thank you for the changes. Jul 23 12:15:35 koen: I'm going to NMU libsexy 0.1.8, you have ten seconds to object Jul 23 12:15:38 pH5: Are you indeed using anthy or just compiling the packages? Jul 23 12:17:55 re Jul 23 12:18:49 i'm still triing to get openzaurus to build. works fine but i wonder where i define what module gets packed into which ipkg when rebuilding the kernel? Jul 23 12:19:39 "rebuild the kernel" == bitbake linux-openzaurus-2.6.16 Jul 23 12:21:48 Laibsch: Currently I'm just compiling, it's several months since I have used it myself. Jul 23 12:22:43 pH5: You should start using your Z (?) again for productive things related to nihongo. I think it is a great tool. Jul 23 12:23:15 I have gotten all my stuff back regarding lookup of words (wadoku, kojien, etc.) Jul 23 12:23:43 I need to put my japanese wikipedia back on the CF card. Jul 23 12:25:10 Laibsch: I got a little sidetracked by my htc phone which still needs a lot of kernel love. Jul 23 12:32:57 ok can't build gcc :( Jul 23 12:45:11 man I love those easy to understand error message : /root/slug/openslug/tmp/work/gcc-cross-4.1.1-r5/temp/run.do_patchcleancmd.17978: line 500: syntax error near unexpected token `(' Jul 23 12:51:43 chouimat: I messed up somewhere Jul 23 12:51:57 chouimat: I'm trying to find out what's wrong Jul 23 12:52:54 pH5: of course you mention that when I'm away ;) Jul 23 12:52:59 koen|away: thanks ... it's a very nice message btw ... which line 500 ;) Jul 23 13:05:13 Laibsch: do you know why the oz-ja people don't cooperate with the oz people? Jul 23 13:05:33 do they not ? Jul 23 13:06:13 koen: I am trying to make things more fluid. I think they are also trying. Jul 23 13:06:16 koen usually jp people are scared of western people ... at least in OSS Jul 23 13:06:21 Any particular complaint? Jul 23 13:06:41 chouimat: I think that is one very valid point. Jul 23 13:06:42 no complaint, it seems like such a wasted effort Jul 23 13:06:52 That is why I said to never yell at them. Jul 23 13:07:00 koen: what wasted effort? Jul 23 13:07:16 I need the Japanese stuff, so for me it is certainly no wasted effort. Jul 23 13:07:46 Laibsch: I know ... cultural difference is huge here ... Jul 23 13:08:04 koen: The general answer is fear, fear of being exposed among others. Jul 23 13:08:15 Laibsch: duplicated work is wasted effort. That's why it is good that you're trying to coordinate. Jul 23 13:08:21 chouimat: What is you background on this? Jul 23 13:09:15 pH5: That is certainly true and the reason I asked. If you find any kind of duplication anywhere, let me know. That is exactly why I try to get stuff into OE quickly so the ppl in .jp realize that Jul 23 13:09:25 Laibsch: I had to work with some japaneses and chineses ... and sometime I would say or do something and they get pissed off ... that kind of thing ... Jul 23 13:09:57 a) we work on this and b) there is no need to "fork" (which I assume to them is not a bad thing as they do not see the downside.) Jul 23 13:10:07 chouimat: I see. Jul 23 13:11:27 Laibsch: like giving a gift to one guy for his birthday but he didn't accepted it because the gift was beyond the guy status ... we want to give him this because we appreciated the guy work more than his japanese manager ... the manager was really pissed off too :) Jul 23 13:11:53 So, please let me know if you find a Japanese person publishing OE-based stuff that is not in OE itself. openzaurus-ja is also publishing binaries without source. They are just unaware of what FOSS really is. They are still at an early stage of the learning curve. Jul 23 13:12:09 chouimat: uhoh! Jul 23 13:12:12 ;-) Jul 23 13:12:27 koen, i think you misinterpreted my request - the problem appears to be that getting started doesn't mention .oz354x (but the sample local conf file does and every time somebody inthe channel says theyre using .dev, coredump or someone else tells them not to use it) Jul 23 13:12:56 poushag: it's the OE gettingstarted, not the OZ one Jul 23 13:14:07 Laibsch: so now I try to avoid to some extend to work with them ... ok if the guy is born here this is different Jul 23 13:17:25 hey zecke Jul 23 13:17:31 RP: may I ask why you renamed openembedded to meta? Jul 23 13:20:24 zecke: isn't removing COPYING.MIT illegal? Jul 23 13:21:08 koen: well, not if it is added to the files directly ;) Jul 23 13:21:12 koen: but calm down :) Jul 23 13:21:14 zecke: It was thought easier to understand by mallum Jul 23 13:22:18 PR: re, I've now build my first .bb file, and got nice .ipk Jul 23 13:22:18 RP: I kind of liked the branding Jul 23 13:22:56 RP: and I think one should readd a licensing file Jul 23 13:23:03 RP: and is oh-image still present? Jul 23 13:23:16 zecke: Its now known as oh-image-pda Jul 23 13:23:27 (see the images directory under packages) Jul 23 13:23:28 and oh-image-pda-sdk? Jul 23 13:23:37 hello everyone Jul 23 13:23:50 RP: what would you like to get built automatically? Jul 23 13:24:02 zecke: oh-image-sdk until it turns into a DISTRO_TYPE Jul 23 13:24:18 zecke: I'm happy with oh-image-pda or the sdk Jul 23 13:24:35 zecke: I wasn't aware a licence file had gone missing? Jul 23 13:24:57 RP: copying.mit in the OE topdir Jul 23 13:24:58 RP: I don't know if it was missing from the beginning Jul 23 13:25:00 PR: I build openvpn. there the files to pack seams to be defined in do_install_append. but I would need to pack some additional kernel modules. where do i define which module gets packed into what ipk? Jul 23 13:25:29 RP: and I would be happy if o-hand would stick with the MIT license to ease merging back :) Jul 23 13:26:03 zecke: I suspect it was missing from the start (and that wasn't intentional) Jul 23 13:27:02 okay. I'm building oh-image-pda and oh-image-sdk for n770 and qemu again Jul 23 13:27:20 so new monotone stuff works now? Jul 23 13:27:31 Crofton: it should (tm) Jul 23 13:27:40 Crofton: I've done a pull from mtn.oe.org.. seems to work Jul 23 13:27:52 good Jul 23 13:27:53 And bloody hell.. it's fast. Jul 23 13:28:00 I'll try this afternoon Jul 23 13:28:05 44 revs in about a second or two.. Jul 23 13:28:07 zecke: Thanks and sorry about the rearranging - not something that should happen again any time soon Jul 23 13:28:27 RP: don't worry. I thought the tinderbox was broken ;) Jul 23 13:29:39 * koen whistles innocently Jul 23 13:29:48 03koen 07org.oe.dev * rc3c1ba81... 10/ (1 packages/gcc/gcc_4.1.1.bb): gcc 4.1.1: fix build troubles Jul 23 13:30:09 later guys, I'm playing some game cube now Jul 23 13:30:26 thanks for getting the db upgraded! Jul 23 13:31:11 03koen 07org.oe.packaged-staging * r6c7c8a36... 10/ (1 COPYING.MIT): add COPYING.MIT Jul 23 13:31:21 Crofton: if you have private branches, commits I would be glad if you could test my script Jul 23 13:31:24 koen: haha!!! Jul 23 13:31:27 koen: ;-) Jul 23 13:31:42 :p Jul 23 13:31:45 koen: we say "don't throw rocks if you sit in a house made of glas" Jul 23 13:31:46 I do, but I might just rebuild the brnach Jul 23 13:31:57 since I don't have much in it Jul 23 13:32:04 Crofton: make a backup, test my script ;) Jul 23 13:32:11 Crofton: if it fails just rebuild your branch Jul 23 13:32:16 and I am not sure how well I kept it updated due to monotone headaches Jul 23 13:32:16 RP: You wouldn't happen to know if OZ builds at all at present, in 354x or dev, would you? Jul 23 13:32:22 ok Jul 23 13:32:26 true Jul 23 13:32:45 NAbyss: 354x should work, dev works if you have some magic local.conf entries to use gcc 3.4.4 Jul 23 13:32:46 NAbyss: what is the question? Jul 23 13:33:01 NAbyss: http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenZaurus-354x Jul 23 13:33:09 zecke: I'm getting build failures with uicmoc-native in oz354x Jul 23 13:33:11 NAbyss: some builds of oz354x on my debian stable Jul 23 13:33:30 NAbyss: pastebin it please Jul 23 13:33:35 zecke: Shall do Jul 23 13:33:38 hello everyone Jul 23 13:35:00 koen, zecke: I added the licence files to poky as well Jul 23 13:35:35 hey fijam Jul 23 13:36:29 thanks Jul 23 13:38:46 RP: the split from meta/ into tasks/ and images/ is nice Jul 23 13:39:18 zecke: Shall have the URL soonish.. pastebin's bloody slow today Jul 23 13:39:19 koen: I have no objection if OE wants to follow that :) Jul 23 13:39:50 koen, this is the oe channel - thats why i mentioned it here - and im being confused while trying to get started using oe the develop for oz Jul 23 13:40:51 is there a bb that builds (and packs) an external kernel module? Jul 23 13:41:10 in order to help me getting startet with building truecrypt Jul 23 13:42:24 ierk: hostap-modules for instance Jul 23 13:42:28 or fuse-modules Jul 23 13:43:08 koen: thx Jul 23 13:43:40 now all i need to be tought is how to pack additional modules build by the kernel itself Jul 23 13:44:15 ierk: the kernel.bbclass should already package all modules it builds Jul 23 13:45:09 zecke: http://i3.tucuxi.org/oetemp - pastebin is on the fritz. Jul 23 13:45:38 03koen 07org.oe.dev * rda3bdc9f... 10/ (93 files in 5 dirs): Jul 23 13:45:38 Split out *image* and *task* from packages/meta into their own subdirs. Jul 23 13:45:38 Inspired by http://svn.o-hand.com/view/poky?rev=531&view=rev Jul 23 13:45:52 RP: there you go Jul 23 13:46:06 koen: Nice, thanks :) Jul 23 13:46:56 and I updated monotone on oe.org to 0.28 Jul 23 13:47:26 should avoid a few deadlocks with lots of clients Jul 23 13:47:34 Does 0.28 bring many benefits? Jul 23 13:47:47 RP: yes, the cherrypicker Jul 23 13:48:13 pluck [-r FROM] -r TO [PATH...] Jul 23 13:48:13 Apply changes made at arbitrary places in history to current workspace. Jul 23 13:50:14 * RP wonders why mtn hangs when I ask it to pull the oz354x branch :-/ Jul 23 13:55:30 RP: i just tried oz354x and it works for me Jul 23 13:55:46 Greg2: I think my database is somehow broken... Jul 23 13:56:01 koen: there is no fuse-modules in oz354x Jul 23 13:56:14 it is in .dev Jul 23 13:56:20 which is what I use :) Jul 23 14:07:06 "Its development began in the late 1980's and it was a research project intended to address a variety of system scalability issues that the UNIX and LINUX kernels don't do particularly well," Jul 23 14:07:14 linux in 1980? Jul 23 14:07:29 plan9 must be really good if they had linux in 1980 :) Jul 23 14:12:16 RP: which version of mtn? Jul 23 14:14:12 zecke: Did that URL work for you? Jul 23 14:15:03 03pH5 07org.oe.dev * rca78f32f... 10/ (1 packages/libnss-mdns/libnss-mdns_0.8.bb): libnss-mdns: add 0.8 Jul 23 14:15:09 03pH5 07org.oe.dev * rc3d1176e... 10/ (1 packages/libsexy/libsexy_0.1.8.bb): libsexy: add 0.1.8 Jul 23 14:15:15 03pH5 07org.oe.dev * rb7862d70... 10/ (1 packages/xlibs/libxfont_X11R7.1-1.2.0.bb): libxfont: add 1.2.0 for ft 2.2 compatibility Jul 23 14:15:20 03pH5 07org.oe.dev * r206b9c34... 10/ (1 packages/notification-daemon/notification-daemon_0.3.5.bb): notification-daemon: add 0.3.5 Jul 23 14:15:26 03pH5 07org.oe.dev * r6e33e6f5... 10/ (13 files in 5 dirs): dbus: add 0.90, dbus-glib and dbus-python bindings 0.70 Jul 23 14:15:31 03pH5 07org.oe.dev * rba784a23... 10/ (1 packages/avahi/avahi_0.6.12.bb): avahi: add bugfix release 0.6.12 Jul 23 14:20:47 http://www.linuxjournal.com/articles/lj/0139/8177/8177f13.png Jul 23 14:20:54 now that's a cool lunchbox Jul 23 14:23:57 zecke: 0.27 Jul 23 14:35:20 RP: there was a bug with mtn hanging on sync Jul 23 14:35:33 RP: might try mtn 0.26 or 0.28 Jul 23 14:36:27 zecke: Ah, ok. I'll try 0.28 Jul 23 14:37:08 koen: got an article to your image? Jul 23 14:37:25 ierk: http://www.linuxjournal.com/article/8177 Jul 23 14:37:56 how do i copy a file from my package's dir to the build dir? Jul 23 14:38:17 adding a "file://filename" to SRC_URI did not do the trick Jul 23 14:39:40 koen I'm building one ... ok without the case :) Jul 23 14:42:30 ierk: http://www.linuxjournal.com/article/8177 Jul 23 14:58:34 re Jul 23 14:59:09 ierk: copy it from ${WORKDIR} to ${S} in one task Jul 23 14:59:21 ierk: e.g. do_configure_prepend() { copy stuff } Jul 23 15:08:50 03pH5 07org.oe.dev * r7e3ebf6d... 10/ (1 packages/linux/handhelds-pxa-2.6_cvs.bb): handhelds-pxa-2.6_cvs: deploy before do_package instead of do_build. this makes sure rm_work isn't called before deploy Jul 23 15:17:02 pH5: tomoe builds fine here but anthy-native fails. Jul 23 15:23:57 Laibsch: can you pastebin the log? Jul 23 15:26:46 hey Jul 23 15:27:03 have any of you guys used the oe klibc stuff? Jul 23 15:27:05 pH5: did you manage to boot EABI userland yet? Jul 23 15:27:15 tkp: it doesn't compile for me Jul 23 15:27:24 koen: it compiles fine for me Jul 23 15:28:26 I'm just a little confused as to how one might use it Jul 23 15:28:40 it has created a load of binaries in tmp/staging/i586-oe-linux/klibc/lib/klibc/i586-oe-linux-bin Jul 23 15:29:08 tkp: I think klibc needs work Jul 23 15:29:09 which I would assume are statically compiled binaries suitable for use in an initramfs (which is my intent) Jul 23 15:29:27 zecke: have you used it? Jul 23 15:29:35 koen: for quite some time, but I didn't try busybox 1.2.0 yet. Jul 23 15:29:37 tkp: no, it is on my agenda though Jul 23 15:29:57 koen: there is a gpe-image build in progress, though. Jul 23 15:30:08 when I try to use one of these binaries, I get the message "No such file or directory" Jul 23 15:30:34 tkp: doesn't ring a bell at all Jul 23 15:30:48 tkp: what does strace say? (do not paste here) Jul 23 15:31:04 zecke: have you sucessfully used any of the created binaries? Jul 23 15:31:10 ok... hold on Jul 23 15:31:16 not built one yet Jul 23 15:31:29 I have just read bits of the source and bits of the disucssion Jul 23 15:31:55 http://rafb.net/paste/results/5omRWJ64.html Jul 23 15:32:10 tkp: I want you to look into the log Jul 23 15:32:24 zecke: which log? Jul 23 15:32:25 tkp: "No such file or directory" could come from a failing open/exec Jul 23 15:32:35 tkp: the one you have pasted :) Jul 23 15:32:41 take it as Free training Jul 23 15:32:46 ;) Jul 23 15:33:20 I've not Jul 23 15:33:31 really had any experience with strace Jul 23 15:36:14 tkp: form a theory why your binary is not working as expected (if you have some sort of expectations) Jul 23 15:36:41 tkp: well do a ls -la Jul 23 15:36:58 and these 'files' are symlinks pointing to nowhere? Jul 23 15:37:11 no, they are not symlinks Jul 23 15:37:33 tkp: see, your strace log has only 12 lines, have fun investigating Jul 23 15:38:10 also man execve could give some clue about the error you get Jul 23 15:39:23 hmm.. odd, I have no man entry for that Jul 23 15:39:31 install the man pages Jul 23 15:40:54 I guess my other option could be to install busybox into the initramfs Jul 23 15:41:08 tkp: you want to give up now? Jul 23 15:41:20 zecke: no... but I'm running out of time Jul 23 15:42:10 * tkp curses the one who invented deadlines Jul 23 15:44:27 then use busybox :) Jul 23 15:45:42 koen: I will hack tinderbox this week :) Jul 23 15:45:52 zecke: I suspected that :) Jul 23 15:46:06 zecke: so I started with the infrastructure :) Jul 23 15:46:27 and if time permits I'll try do de-uglify it Jul 23 15:46:37 3d rotating OE logos and stuff Jul 23 15:48:45 * koen is getting to close to having a working fortran Jul 23 15:48:51 thanks to Lardman Jul 23 15:49:13 it is too hot for doing anything Jul 23 15:49:46 Laibsch: Tested file: /home/oe-build/oe/org.openembedded.dev/packages/anthy/anthy-native_7811.bb Jul 23 15:49:48 * chouimat is freezing here Jul 23 15:49:49 Test Result: False Jul 23 15:49:52 Test Comment:Native package is RDEPENDS_anthy'ing on non native package 'libanthy0' Jul 23 15:50:46 zecke: What is this test thing? Jul 23 15:51:00 bittest using depends_checker Jul 23 15:51:18 http://rafb.net/paste/results/2tIWnA10.html Jul 23 15:51:53 is there a way to install an existing .ipk file somewhere? Jul 23 15:52:10 obviously... but what is it?! Jul 23 15:52:11 tkp: like one does with the rootfs? Jul 23 15:52:12 tkp: ipkg install foo.ipk Jul 23 15:52:25 zecke: I guess I screwed up with anthy after all. Jul 23 15:52:29 but I guess you want the '-o' option Jul 23 15:52:41 -o lets you specify a location? Jul 23 15:52:51 yes (offline root) Jul 23 15:53:06 pH5: could be a false positive as well Jul 23 15:53:06 or '-d' + dest in ipkg.conf Jul 23 15:53:37 hmm, ipkg: command not found Jul 23 15:53:39 zecke: the problem is anthy-native include anthy and forgets to clear out PACKAGES Jul 23 15:54:26 oh hold on... rootfs_ipk.bbclass does ipkg-cl install Jul 23 15:55:04 ouch Jul 23 16:03:05 03pH5 07org.oe.dev * r32037b17... 10/ (1 packages/anthy/anthy-native_7811.bb): anthy-native-7811: clear PACKAGES Jul 23 16:08:07 pH5: pulling right now. I will let you know the result. Jul 23 16:08:35 Is mtn faster than monotone for subsequent pulls? My impression is "not really", or is it? Jul 23 16:09:35 Laibsch: ±10x faster Jul 23 16:10:01 Laibsch: which version of mtn? Jul 23 16:10:02 Why does gpsd depend on dbus 0.23 specifically? Jul 23 16:11:07 zecke: 0.28. It was just my feeling. It was not interactive before and is not now. But again, this is my router with a Via C3@600MHz. Jul 23 16:11:21 RP: afaik we can't handle versioned depends Jul 23 16:11:57 koen: I didn't think we could either but it has a very specific DEPENDS line... Jul 23 16:13:28 just discard it :) Jul 23 16:17:20 reading an old docbook by kergoth we should be able to handle depends like glibc-2.3 Jul 23 16:17:51 zecke: Right, this is how this one supposedly works I guess Jul 23 16:18:09 I have no intention of using dbus 0.23 with it though... Jul 23 16:19:34 hmm Jul 23 16:19:43 "I don't know what todo with myself" :} Jul 23 16:19:50 I think I will hack on the bitbake-c parser Jul 23 16:20:03 any objections if -c clean would remove .ipk files as well? Jul 23 16:21:21 native and non-native? Jul 23 16:21:44 zecke: the ones that go with the recipe Jul 23 16:23:45 koen: why would that be needeD? Jul 23 16:23:54 koen: e.g. if you clean the package, the stamp will be removed Jul 23 16:24:05 koen: and BitBakes DEPENDS handling should recreate the package? Jul 23 16:25:06 zecke: by that reasoning we can leave $WORKDIR intact as well Jul 23 16:29:36 so you want to remove packages to conserve storage? Jul 23 16:29:39 then go ahead Jul 23 16:30:04 if two bb files create the same package, we have a bug... (I guess) Jul 23 16:30:17 no, I want clean to really *clean* Jul 23 16:32:14 well :) Jul 23 16:33:30 What's the best way in oe to work around broken downloads? I had added the tar.gz manually in the past, but don't find a tar.gz for tslib Jul 23 16:33:32 http://pastebin.com/754920 Jul 23 16:33:56 stefan_schmidt: I suspect your cvs install is broken Jul 23 16:35:00 I can try it manually, if you like. But i oe downloads via cvs before today. Jul 23 16:35:32 koen: This time i'm at home, so no network admin block ports. ;) Jul 23 16:37:59 RP, koen: gconf-dbus svn20060119 doesn't compile against dbus 0.90, svn20060723 does. Jul 23 16:38:17 stefan_schmidt: # Jul 23 16:38:17 cvs [checkout aborted]: connect to pubcvs.arm.linux.org.uk(212.18.232.187):2401 failed: Connection timed out Jul 23 16:39:47 stefan_schmidt: try fixing that first :) Jul 23 16:41:19 koen: What? I can't follow you. Jul 23 16:41:57 pH5: add a new snapshot? Jul 23 16:42:31 stefan_schmidt: OE can't help it if a) your cvs install is broken or b) upstream is broken Jul 23 16:42:59 * koen bets it's b) Jul 23 16:43:51 koen: there is no snapshot. it's the _svn.bb locked by sane-srcdates.inc Jul 23 16:44:49 pH5: override it in angstrom-2006.9.conf or add a new snapshot Jul 23 16:45:13 koen: Another cvs update works fine, so it must be b). Anyway, i thought CVS_TARBALL_STASH was there to help out of this problem? Jul 23 16:45:20 koen: ok Jul 23 16:45:44 stefan_schmidt: sure, but two issues. The default URL does not update often Jul 23 16:46:04 stefan_schmidt: and one should really fixate tslib to a specific date Jul 23 16:46:24 oen@bitbake:~/OE/monotone/org.openembedded.dev/packages/gcc$ ls ../tslib/ Jul 23 16:46:24 tslib tslib-maemo tslib-maemo_0.0.1-15.bb tslib_20060703.bb Jul 23 16:46:34 zecke: your a few weeks behind ;) Jul 23 16:46:57 koen: oh treke found the time to help us out? Jul 23 16:47:12 no, mickeyl pinned tslib Jul 23 16:47:58 about three weeks ago Jul 23 16:48:32 zecke: Is there another URL, which works better? Of course you would like to have a pinned tslib version, that is what cvs snapshots for. Jul 23 16:49:07 stefan_schmidt: not that I know of Jul 23 16:49:15 zecke: ok Jul 23 16:49:25 but you are free to provide one Jul 23 16:49:36 * stefan_schmidt waits for the point he can test gpe on his ezx device. Jul 23 16:49:42 http://ewi546.ewi.utwente.nl/tmp/tslib_pubcvs.arm.linux.org.uk__20060703.tar.gz Jul 23 16:49:54 koen: thanks Jul 23 16:50:05 stefan_schmidt: what other source was missing? Jul 23 16:50:30 koen: It seems that openzaurus-ja publishes the bb files at http://cvs.sourceforge.jp/cgi-bin/viewcvs.cgi/openzaurus-ja/bbfile/packages-ja/ I will try to get some of that stuff into OE and encourage openzaurus-ja to upload their stuff directly. Jul 23 16:50:36 zecke: Nothing yet. Will the on the next run. Jul 23 16:51:43 pH5: anthy-native still fails to build but the log is big and I have trouble paste'ing it. Can I PM you? Jul 23 16:51:43 zecke: What is needed for such a mirror? There are some scripts ready to do the dirty work? Jul 23 16:52:07 Laibsch: sure Jul 23 16:52:10 stefan_schmidt: "jein" Jul 23 16:52:16 stefan_schmidt: the sources are in the diretory you specified. Jul 23 16:52:20 Laibsch: and tell them to remove that ridiculous branding Jul 23 16:52:29 stefan_schmidt: bitbake -cfetch world should do the trick but there might be a bitbake bug Jul 23 16:52:36 koen: That is yelling at them. Jul 23 16:52:42 Laibsch: it's nonsense to prefix everything with 'ozja' Jul 23 16:52:49 koen: and why ridiculous brading? Jul 23 16:53:01 Ok, I see. Jul 23 16:53:10 Laibsch: see http://cvs.sourceforge.jp/cgi-bin/viewcvs.cgi/openzaurus-ja/bbfile/packages-ja/libeb/ Jul 23 16:53:27 and inside that .bb they have to do weird stuff to work around that very branding Jul 23 16:53:34 hence 'ridicilous' Jul 23 16:53:39 koen: that problem would be solved automatically if they pushed their stuff upstream instead Jul 23 16:53:42 (and I can't spell) Jul 23 16:53:49 :) Jul 23 16:54:03 I will try to get them to do that. Jul 23 16:54:12 I guess it would certainly be the best solution. Jul 23 16:54:14 zecke: 20 GB disk is neded for this, right? Jul 23 16:54:26 stefan_schmidt: ~8GB Jul 23 16:54:39 stefan_schmidt: but the question is when to remove snapshots? Jul 23 16:54:46 * stefan_schmidt thought the wiki talks about 20GB Jul 23 16:55:27 koen: would you think anyone minds if I adapt the darwinports distfiles naming? Jul 23 16:55:31 stefan_schmidt: for builds Jul 23 16:55:32 Sorry, 20GB for a bitbake world, not only fetch Jul 23 16:55:36 pH5: Finally got through: http://oe.pastebin.com/754924 Jul 23 16:56:49 zecke: Isn't it the deciscion of the package maintainer, when to remove a snapshot? Jul 23 16:57:28 stefan_schmidt: snapshot as in floating cvs/svn/git SRCDATE Jul 23 16:57:45 stefan_schmidt: so there is no 'maintainer' decision, it all depends on the user running bitbake Jul 23 16:58:06 stefan_schmidt: and well some distributions will pin it to a date so removing the oldest might not be the most clever solution Jul 23 16:58:42 zecke: hmm, you're right. Bigger problems i guess. Jul 23 16:59:20 stefan_schmidt: that part of the floating SRCDATE is flawed :) Jul 23 17:00:24 zecke: What's about doing cvs snapshots every day for a month? Jul 23 17:00:43 zecke: Distro which pins so old version should use a release. :) Jul 23 17:00:45 stefan_schmidt: and then? remove the files with the least hits? Jul 23 17:00:56 stefan_schmidt: e.g. tslib has no release Jul 23 17:01:02 that breaks any old builds someoen might need Jul 23 17:01:32 not that they would currently work anyway ... Jul 23 17:02:23 * zecke will get him another pizza box of the rack assigned to do OE builds Jul 23 17:03:02 * emte sends zecke a pepperoni pizza instead Jul 23 17:03:29 emte: oh well. I kind of like the Xeon Systems Jul 23 17:03:50 too hard on the teeth Jul 23 17:04:01 and bad indigestion Jul 23 17:05:47 whats the smallest xserver availible for oe? Jul 23 17:06:55 zecke: hmm, we could perhaps extend the periode to and flaf or even full year, with some symlink magic, if cvs checkouts have not changed. Jul 23 17:08:16 jkp: the one with 'tiny' in the name ;) Jul 23 17:08:38 opie-tinykate? Jul 23 17:08:46 nope Jul 23 17:08:51 (that is a text editor) Jul 23 17:08:55 kdrive/tinyY Jul 23 17:09:06 i dont see thgat in my checkout Jul 23 17:10:12 definately no package with the word tiny in it that is related to x that i can see :/ Jul 23 17:10:17 kdrive? Jul 23 17:10:38 not at the top level Jul 23 17:10:42 xserver/ Jul 23 17:10:46 ok Jul 23 17:10:59 got it Jul 23 17:11:10 cool - any ideas how small it goes? Jul 23 17:11:44 you'd want diet-x11 as xlib to go with it Jul 23 17:12:14 424K xserver-kdrive-fbdev_0.0+cvs20060312-r11_armv5te.ipk Jul 23 17:12:23 cool :) Jul 23 17:12:33 does that include the lib? Jul 23 17:13:04 also one more thing, small issue - i need to avoid any flickering between starting X and it taking over the screen, and the splash screen...doable? Jul 23 17:15:00 it is, the maemo people have a patch for that Jul 23 17:15:52 koen: cool Jul 23 17:19:01 pH5: if that new images works upto the serial prompt, could you put it online? Jul 23 17:19:51 03zecke123 * r550 10bitbake_qa/ (5 files in 5 dirs): bitbake_qa: Ignore *.pyc and other files Jul 23 17:20:02 oops, why is that working again? Jul 23 17:20:13 zecke: because kergoth fixed it Jul 23 17:20:23 bad kergoth ;) Jul 23 17:20:24 we all like it :) Jul 23 17:22:07 ERROR: The following packages conflict with bluez-utils-doc: bluez-utils-nodbus Jul 23 17:22:09 wtf? Jul 23 17:22:23 RPROVIDES gone wrong? Jul 23 17:26:22 zecke: it does send the commits to the wrong project, openembedded, instead of bitbake Jul 23 17:26:43 zecke: see http://cia.navi.cx/stats/author/zecke123 and Jul 23 17:28:39 03zecke123 * r551 10bitbake_qa/conf/bitbake.conf: Jul 23 17:28:39 bitbake_qa: Minor updates on the bitbake.conf Jul 23 17:28:39 Update GIT_DIR and mirrors. These changes were taken Jul 23 17:28:39 from the bitbake.conf of OE. Jul 23 17:29:28 koen: should I fix the binutils issue now? or did you fix it? Jul 23 17:29:57 go ahead Jul 23 17:30:07 koen: 6 thousand more revs ;) Jul 23 17:33:02 koen: what is the package that builds xserver-kdrive-fbdev? its telling me nothign provides it :/ Jul 23 17:33:12 ls Jul 23 17:33:36 ok, got it Jul 23 17:33:46 i guess i need to provide a preferred provider Jul 23 17:35:29 03zecke123 * r552 10bitbake/lib/bb/parse/parse_c/bitbakec.pyx: bitbake/lib/parser/parse_c: Avoid some segfaults Jul 23 17:37:36 RP: could you take the tinderclient.bbclass from .dev and put it into poky? Jul 23 17:38:07 zecke: ok Jul 23 17:39:25 zecke: done Jul 23 17:45:37 so kergoth is back to bitbake development :) Jul 23 17:45:48 zecke: yes Jul 23 17:46:39 good Jul 23 17:51:06 gpsdrive is starting to annoy me now. Hardcoded references to /usr/include, broken configure.ac :-/ Jul 23 17:51:25 RP: pokify maemo-mapper adn put it back in OE Jul 23 17:51:34 s/adn/and/ Jul 23 17:51:52 koen: Its too limited for what I want to do :-( Jul 23 17:52:05 koen: I already have it running on the 770... Jul 23 17:52:10 I wouldn't try to waste time on gpsdrive Jul 23 17:52:58 good evening ! Jul 23 17:53:32 hey cyrilRomain Jul 23 17:53:43 hey koen! Jul 23 17:53:52 * cyrilRomain missed OE for too long .... :) Jul 23 17:54:28 koen: how is your SoC going ? Jul 23 17:54:43 pretty good Jul 23 17:56:55 cyrilRomain: http://dominion.kabel.utwente.nl/koen/cms/progress-with-my-summer-of-code Jul 23 17:58:11 koen: great :) Jul 23 18:00:48 03koen 07org.oe.dev * rc1ab9b83... 10/ (1 packages/angstrom/task-angstrom.bb): task-angstrom: s/diet-x11/libx11/ Jul 23 18:03:25 ah Jul 23 18:03:26 NOTE: package angstrom-gpe-image-1.0-r1: task do_rootfs: started Jul 23 18:04:05 so your running from monotone 0.27 now right? Jul 23 18:04:26 anyone know where http://www.oesources.org/source/current/tslib_pubcvs.arm.linux.org.uk__20060703.tar.gz has gone? Jul 23 18:04:40 it was never there Jul 23 18:04:40 was it ever here? Jul 23 18:04:53 koen: well the bb file wants it Jul 23 18:04:56 (current one) Jul 23 18:05:00 no it doesn't Jul 23 18:05:07 it wants the cvs repo Jul 23 18:05:11 which is down Jul 23 18:05:15 :/ Jul 23 18:05:21 and it tries to fall back to the tarball Jul 23 18:05:34 put http://ewi546.ewi.utwente.nl/tmp/tslib_pubcvs.arm.linux.org.uk__20060703.tar.gz in your DL_DIR Jul 23 18:05:40 koen: It now appears to run on the 760. The screen is totally the wrong size for it though :-/ Jul 23 18:05:48 koen: thanks Jul 23 18:06:00 RP: remove the -p from the .desktop Jul 23 18:06:13 RP: but gpsdrive is &$*(@)$ broken in every way Jul 23 18:06:45 it uses gpsd in the wrong way, every bit of code if duplicated with #ifdef pda, their projections are wrong Jul 23 18:06:48 etc... Jul 23 18:09:20 RP: I replaced gpsdrive with a bunch of shellscript for my trip to florida, no gui but at leat it worked :) Jul 23 18:09:21 koen: The only sane program is maemo mapper then? Jul 23 18:09:36 maemo-mapper is heading in the right direction Jul 23 18:09:43 koen: What did those shell scripts do? Jul 23 18:10:16 log lots of stuff to a .gpx file that I could run through gpsphotolinking Jul 23 18:10:37 * koen tries to find the scripts Jul 23 18:10:54 koen: Did it include logging speed information? Jul 23 18:11:06 yes Jul 23 18:11:30 * RP would be interested as that combined with maemo-mapper might work for what I want :) Jul 23 18:11:47 ask gpsd to output 'v' Jul 23 18:11:48 gna, where are the kernel-module-$foo packages defined? Jul 23 18:11:56 'v' is in knots iirc Jul 23 18:12:04 * koen tries to remember Jul 23 18:12:15 it could be that nmea reports knots and gpsd km/h Jul 23 18:12:51 koen: NEMA is marine based so that sounds likely. I'd want miles/h of course ;-) Jul 23 18:13:11 gah Jul 23 18:13:20 RP: nautic miles or us miles Jul 23 18:13:27 where is my scripts/ subfolder Jul 23 18:13:37 ierk: uk ones ;-) Jul 23 18:14:11 * ierk always knew that the miles and inces thing was fishy Jul 23 18:14:21 koen: I'm being called away - back later. If you did find some scripts, I'd be interested though. Of course the 770/maemo-mapper doesn't use gpsd but I suspect I could change that :) Jul 23 18:14:50 why didn't the world switch over to the metric system yet? and while at it the iso date/time format Jul 23 18:16:38 03justinp 07org.oe.oz354x * r0a725bec... 10/ (4 files in 2 dirs): build-oz-jpatrin: a buildscript that works for newest oz354x using the interactive shell Jul 23 18:37:10 ade|home: top gear time :) Jul 23 18:44:10 whats the deal with busybox_1.2.0? Jul 23 18:44:15 is it working? Jul 23 18:44:45 I just updated my checkout, but running bitbakke busybox seems to default to building 1.01 Jul 23 18:45:00 koen: do we only need to install xserver-kdrive-fbdev_0.0+cvs20060312-r11_i586.ipk? Jul 23 18:45:18 (which built without any problems :) ) Jul 23 18:49:18 zecke? Jul 23 18:49:22 yes? Jul 23 18:49:30 where are your migration scripts? Jul 23 18:49:38 Crofton: will upload them Jul 23 18:49:41 ok Jul 23 18:49:56 Can I go ahead and upgrade monotone? Jul 23 18:50:08 hmmm Jul 23 18:50:10 you need both binaries in your path ;) Jul 23 18:50:21 they jsut release 0.28 ... Jul 23 18:50:26 is that version ok? Jul 23 18:50:31 hmmm Jul 23 18:50:38 I have not tested it Jul 23 18:50:40 it should be fine Jul 23 18:50:46 hi all Jul 23 18:50:53 rename old monotone to ? Jul 23 18:51:04 Crofton: old monotone: monotone new monotone: mtn Jul 23 18:51:13 the bb file for busybox 1.2.0 has DEFAULT_PREFERENCE = "-1" Jul 23 18:51:21 which stops it being picked Jul 23 18:51:31 any ideas why that's in there? Jul 23 18:52:06 ah Jul 23 18:52:23 so when I build new monotone it calls itself mtn? Jul 23 18:52:37 CoreDump|afk: yes Jul 23 18:52:39 Crofton: yes Jul 23 18:52:56 this will confuse me :) Jul 23 18:53:19 whom should i mail updated .bb files? the maintainer? Jul 23 18:53:58 bugzilla Jul 23 18:55:17 zecke: bugs.openembedded.org? Jul 23 18:55:46 anyone got a login for me? ;-) Jul 23 18:55:55 create one Jul 23 19:01:14 Crofton: my hh.org account doesn't work for some reason Jul 23 19:03:02 he zecke Jul 23 19:05:17 heh Jul 23 19:05:39 well, I may create a new db, but I will keep the one with te private branch around Jul 23 19:05:54 hmm Jul 23 19:06:03 Crofton: oh well :) Jul 23 19:06:07 I am sure it is messy, because of the pain pulling changes from dev Jul 23 19:06:25 I am not certain I handled merges properly Jul 23 19:06:31 ah okay Jul 23 19:06:58 I am going to be very unhappy if the new mtn does not merge across branches better .... Jul 23 19:07:09 ah ok I think I see why now.. Jul 23 19:07:11 well, if you want to take this oppurtunity to start over again do it Jul 23 19:07:17 yep Jul 23 19:07:17 busybox 1.2.0 does not yet work :( Jul 23 19:07:35 otherwise you could do the migrating by hand Jul 23 19:08:10 I don't think it will be painful enough to really need to automate Jul 23 19:09:42 Crofton: okay then make a backup :) Jul 23 19:09:43 How do I create a feed out of my tmp/deploy? IOW, how do I create the Packages file? Jul 23 19:09:50 will do Jul 23 19:09:59 Crofton: mtn --db to_be_changed.db db migrate Jul 23 19:10:02 Crofton: mtn --db to_be_changed.db db rosterify Jul 23 19:10:04 Crofton: wait.... Jul 23 19:11:05 who is working on the new busybox version? Jul 23 19:15:35 ok think is really getting scary now ... Jul 23 19:22:19 basically, when trying to use busybox 1.2.0, it complains that libgcc_s.so.1 is missing Jul 23 19:22:43 copying the offending file in manually seems to have fixed it from what I can tell Jul 23 19:23:25 hmm, when I DEPENDS something in a .bb file, it get's installed in my staging dir, but for some reasons the /usr/share/aclocal files are not installed Jul 23 19:23:44 i'm triing to pack cryptsetup what DEPENDS libgcrypt Jul 23 19:31:08 hmmm Jul 23 19:31:15 install texi2html ... Jul 23 19:31:20 ? Jul 23 19:33:56 ierk: aclocal should be staged? Jul 23 19:40:24 zecke: well, if i do it manualy configure works fine Jul 23 19:40:26 tkp: busybox 1.2.0 has only been tested on angstrom, all other distro's need to update their busybox defconfigs Jul 23 19:41:11 koen: I'm trying to build busybox into an initramfs Jul 23 19:41:21 whhich ideally should be a static version of busybox Jul 23 19:41:33 ftp.handhelds.org connection refused? Jul 23 19:41:52 Crofton: heavily loaded ATM Jul 23 19:41:56 ok Jul 23 19:42:01 the problem being that busybox is attempting to link dynamically to libgcc_s.so.1 (which does not exist on my initramfs image) Jul 23 19:42:16 just trying to build boostrap-image against new db Jul 23 19:42:26 simply copying the library in seems to make things work ok Jul 23 19:44:11 * koen tries 'mtn pluck' Jul 23 19:45:59 03oe 07org.oe.oz354x * ra8efffa3... 10/ (1 classes/base.bbclass): Jul 23 19:45:59 pluck r2e86ad9f702085cdc43db32345744bae3a2268eb from .dev Jul 23 19:45:59 applied changes from 580919464ad15e7c6f626f4777c1683ad8651be3 Jul 23 19:45:59 through 2e86ad9f702085cdc43db32345744bae3a2268eb Jul 23 19:46:18 I did use the wrong key, though Jul 23 19:50:47 is there a variable that contains the build directory? Jul 23 19:51:02 (as in the path to tmp/build) Jul 23 19:51:25 * ierk got cryptsetup to build and install. Jul 23 19:51:43 tkp: see conf/bitbake.conf Jul 23 19:51:45 Crofton: if your connection problem comes when fetching ipkg-native_0.99.163, I found a little typo in SRC_URI: http://pastebin.com/755040 Jul 23 19:51:55 had to copy one file manualy as it did not get staged and am now fighting with getting cryptsetup to work ;-) Jul 23 19:51:59 good nicht all Jul 23 19:52:07 ah Jul 23 19:52:18 Thanks cyrilRomain I'll check that Jul 23 19:53:31 that fixed it Jul 23 19:53:45 is it in bugzilla? Jul 23 19:54:33 koen: thanks... wish I knew about that file sooner :) Jul 23 19:54:42 although there is no BUILD_DIR variable Jul 23 19:56:22 Crofton: not yet. I file the bug + fix right now Jul 23 19:57:13 03c.romain 07org.oe.dev * rdb8a27d3... 10/ (1 packages/ipkg/ipkg-native_0.99.163.bb): ipkg-native .163: fix SRC_URI Jul 23 19:57:25 03zecke123 * r553 10bitbake/lib/bb/parse/parse_c/ (BBHandler.py bitbakec.pyx bitbakescanner.cc bitbakescanner.l): Jul 23 19:57:25 lib/bb/parse/parser_c: Handle includes, requires, cope with our syntax Jul 23 19:57:25 -be able to parse variables like PREFERRED_PROVIDER_virtual/kernel Jul 23 19:57:25 -handle includes and requires, add 's' to %(fn) Jul 23 19:57:27 -be less verbose Jul 23 19:57:29 thanks Jul 23 19:57:38 cyrilRomain: there you go :) Jul 23 19:57:45 ah thanks koen :) Jul 23 19:57:51 * koen hugs mtn commit --author Jul 23 19:58:15 Could not include file nslu2-linksys-kernel_2.4.22.bb Jul 23 19:58:16 Segmentation fault (core dumped) Jul 23 19:58:17 fun Jul 23 19:59:47 rwhitby: the sync algo works best when it has its own .mtn file to work on, save you from locking hell Jul 23 20:00:17 hmm where does it put its coredump Jul 23 20:00:32 should the $pwd Jul 23 20:00:46 right Jul 23 20:03:05 RP: could you send a mail to oe@ highlighting your fancy new tasks? Jul 23 20:04:27 rwhitby: UNSLUNG_KERNEL_EXTRA_SRC_URI ?= <- makes the bitbake C parser crash Jul 23 20:09:27 rwhitby: should this be an empty string? Jul 23 20:14:59 hmm 4 seconds to look at each and every of the 3549 Jul 23 20:15:18 I'm some how confident to get below 15 seconds on initial parsing :} Jul 23 20:15:48 put the files in tmpfs :) Jul 23 20:16:03 do we get that time already? Jul 23 20:16:12 put a big Opteron CPU into your computer Jul 23 20:17:42 zecke: http://www.linuxjournal.com/article/8177 Jul 23 20:18:08 could you make bitbake a 9p client/server? Jul 23 20:18:12 * koen hides Jul 23 20:19:19 Info for gentoo users who want monotone-0.28: you can copy monotone-0.25.ebuild into your portage overlay and rename it to monotone-0.28.ebuild. It worked here with gcc4.1.0 on x86. (i guess it work as well for monotone-0.27) Jul 23 20:19:46 or they could install debian ;) Jul 23 20:19:55 ;) Jul 23 20:21:02 awesome Jul 23 20:24:01 ooooh Jul 23 20:24:01 http://lists.debian.org/debian-arm/2006/07/msg00104.html Jul 23 20:25:48 hi Jul 23 20:32:15 How do I create a feed out of my tmp/deploy? IOW, how do I create the Packages file? Jul 23 20:32:31 bitbake package-index Jul 23 20:38:06 RP: did you read the post on oesf.org regarding poodles sound distortions? Suspending for a few seconds "fixes" it heh Jul 23 20:38:20 what do you do after bitbake meta_gpe? Jul 23 20:40:26 whatever you want? Jul 23 20:40:45 drink rum Jul 23 20:42:01 It was nice to finnaly see it build but I cannot findany documentation on how to use it or was to do next. Jul 23 20:42:23 CoreDump|home: I don't follow the forums that much Jul 23 20:42:41 thought so ;) Jul 23 20:42:51 fora, schmora Jul 23 20:43:05 basically a short apm suspend fixes the sound distortions for whatever reason Jul 23 20:43:28 koen: Thanks for the mail Jul 23 20:43:43 RP: i wish I could find the extended version Jul 23 20:44:42 gpsd rules Jul 23 20:45:16 koen: I'll send a mail to oe@ with a link to that task file and ask for comments Jul 23 20:45:25 ok Jul 23 20:45:46 I'll try to OEify it when I have some time next week Jul 23 20:47:06 hmmm Jul 23 20:47:14 maybe we should have a branch for that Jul 23 20:47:21 org.oe.crazynewstuff Jul 23 20:48:28 I also have http://www.rpsys.net/openzaurus/temp/oe_debug_packages-r0.patch Jul 23 20:48:41 * CoreDump|home thinks .dev is crazy enough ;) Jul 23 20:49:14 oooh Jul 23 20:50:07 I haven't actually tried debugging on an image built with it yet but its doing all the right things seemingly Jul 23 20:50:34 if I had a nickel everytime I heared that Jul 23 20:50:57 you guys already have way too many branches Jul 23 20:51:01 koen: Have some faith ;-) Jul 23 20:51:08 emte: All three of them? Jul 23 20:51:31 three? Jul 23 20:51:35 there are 7 Jul 23 20:51:44 or was the last time i looked Jul 23 20:51:51 IMO a stable release can not built w/ .dev. To much random breakage Jul 23 20:52:01 i agree with CoreDump|home Jul 23 20:52:10 Very true and not what .dev was intended for Jul 23 20:52:16 indeed Jul 23 20:52:17 but it seems too many things are forked too eraly still Jul 23 20:52:22 * CoreDump|home hugs .oz354x Jul 23 20:52:24 early* Jul 23 20:52:32 RP: what is that objcopy magic doing? Jul 23 20:52:38 emte: only one moron has forked Jul 23 20:52:47 * koen shuts up now Jul 23 20:53:01 RP: specially that add-debug-link Jul 23 20:53:02 zecke: Places the debug symbols in a separate file and then links the two Jul 23 20:53:23 holy shit, that is the best thing since HURD ;) Jul 23 20:53:24 how long has oz345x been forked now? Jul 23 20:53:28 7 months>? Jul 23 20:53:44 hmmm not sure really. Jul 23 20:53:45 emte: that's 'branched', not forked Jul 23 20:53:54 correct =) Jul 23 20:54:14 people eerge stuff between the two Jul 23 20:54:27 that is semi true, but there still remain no viably decent tools to move commits across branches Jul 23 20:54:40 emte: mtn pluck? Jul 23 20:54:42 emte: stop living in the past Jul 23 20:54:57 it sounds like 0.29 fixed the issue tho Jul 23 20:55:07 emte: remember what I said about shutting out if you don't know what you're thinking about? Jul 23 20:55:10 emte: that is in the future ;) Jul 23 20:55:20 now is such a moment Jul 23 20:55:28 koen: normally people don't think loud :} Jul 23 20:55:34 koen i remeber but i like to do it to show what an ass you can be Jul 23 20:56:17 * koen prefers ass over ignorant Jul 23 20:56:46 monotone 0.29? Jul 23 20:57:13 koen: relax, get something to drink Jul 23 20:57:20 rum Jul 23 20:57:23 right Jul 23 20:57:28 i concider any of my musings to be simular to a general/occasional user impression nothing more Jul 23 20:57:32 place a picture of emte on your dart board ;) Jul 23 20:57:42 ~users Jul 23 20:57:46 so this is what we get for sharing our unpaid volunteer work with you? complaints and accuses? well done, this clearly supports our motivation to continue working on open source projects. Jul 23 20:57:57 relax guys Jul 23 20:58:02 CoreDump|home: hehe Jul 23 20:58:16 ~professors Jul 23 20:59:48 i am very well aware of the tools availible, including the nice one written by JustinP Jul 23 21:01:23 emte: sure we, Linux in general need better tools. But I think taking our labor force into account we are progressing nicely Jul 23 21:01:35 emte: the monotone folks progress awesome as well Jul 23 21:02:33 yes, monotone is doing very well, and OE is developing very well in certain directions as well Jul 23 21:06:14 koen: Thanks for the hint about package-index Jul 23 21:06:18 * Laibsch compiles away Jul 23 21:12:45 03zecke123 * r554 10bitbake/lib/bb/parse/parse_c/ (7 files): Jul 23 21:12:45 lib/bb/parse/parse_c: Throw Parse Exceptions, crash less often Jul 23 21:12:45 -Make sure TOPDIR is set (do not declare this as an error here) Jul 23 21:12:45 -Do not crash on 'NULL' strings Jul 23 21:12:45 -Throws exceptions properly using 'except -1' from within our Jul 23 21:12:46 cdef. Jul 23 21:13:06 hmm, I think I will turn this off Jul 23 21:13:18 zecke: why? Jul 23 21:13:22 it's very informative Jul 23 21:13:30 I feel observed Jul 23 21:13:40 zecke: Please leave it on, its useful Jul 23 21:13:54 well Jul 23 21:16:36 Is there really good reason that libopie2 depends on bluez-libs? And that the bluetooth-stuff is installed by default? My SL-5500 does not have bluetooth. Jul 23 21:17:09 zecke: yes, UNSLUNG_KERNEL_EXTRA_SRC_URI ?= "" should be fine Jul 23 21:17:12 Laibsch: it has cf Jul 23 21:17:15 take it up with mickeyl when he will return Jul 23 21:17:24 Laibsch: and people - like me - can put a CF card into the slot Jul 23 21:18:05 Well, that does not make it a depend IMHO, it is still optional. Jul 23 21:18:24 Laibsch: It is a build dependeny of libopie2 Jul 23 21:18:31 Laibsch: installing is a different issue Jul 23 21:18:48 Laibsch: using your reasoning one wouldn't install pppd, wireless stuff, serial stuff Jul 23 21:18:50 03koen 07org.oe.dev * rf0347339... 10/ (1 conf/machine/h5xxx.conf): conf/machine/h5xxx.conf: start adding support for a cruftless h5xxx images Jul 23 21:18:55 as these peripherals are optional Jul 23 21:18:55 03koen 07org.oe.dev * rab710aad... 10/ (3 files in 3 dirs): basefiles: add fstab for h5xxx Jul 23 21:19:02 03koen 07org.oe.dev * r6da5a333... 10/ (4 files in 4 dirs): tslib: add support for h5xxx Jul 23 21:20:18 well Jul 23 21:20:36 Where did http://www.oesources.org/source/current/ipkg-0.99.163.tar.gz go? Jul 23 21:20:48 dunno :) Jul 23 21:21:06 And where did ftp.handhels.org go? Jul 23 21:21:16 And where did ftp.handhelds.org go? Jul 23 21:21:26 Laibsch: you have two options. Take it up with mickeyl/hrw and convince him not to include bluetooth b) do it like coredump and create your own customized image Jul 23 21:21:35 Laibsch: it has a load of 25, 25, 25 Jul 23 21:21:53 zecke: I think I will take option b) or remove the stuff after flashing. Jul 23 21:22:24 zecke: wow, 25 is not good. Anybody taking care of the poor thing? Jul 23 21:22:39 france is still asleep Jul 23 21:23:13 ah load is way better Jul 23 21:24:06 I bet it was the poor spam filter :} Jul 23 21:25:02 03koen 07org.oe.dev * r1e46350f... 10/ (3 files in 3 dirs): hh-pxa: add skeleton support for h5xxx Jul 23 21:26:20 Is there a way I can dump into a shell instead of do_compile? I want to work around a menu system, but I can't until I know what I can't see. do_compile() would work also if I could have it output to the console instead of to the log.do_compile. Jul 23 21:26:57 cinix: why not using tail on the log? Jul 23 21:27:06 zecke, I need to interact with it Jul 23 21:27:19 cinix: in which way interact? Jul 23 21:27:38 cinix: e.g. take a look at QtE where it echos 'yes' to configure Jul 23 21:28:02 zecke, right right, once I know what to do. Jul 23 21:28:12 zecke, I need an environment to experiment in Jul 23 21:28:39 cinix: to get the log you would need to change bitbake Jul 23 21:28:42 I think build.py Jul 23 21:28:48 really all I want is it to fetch and leave me with the right PATH= Jul 23 21:28:59 bitbake devshell Jul 23 21:28:59 ah okay. Jul 23 21:29:02 oh well Jul 23 21:29:08 as koen said Jul 23 21:29:15 devshell or interactive bitbake Jul 23 21:30:13 oh interactive is all I need. great thanks Jul 23 21:34:25 'night all Jul 23 21:46:21 03coredump 07org.oe.oz354x * r1ea26db3... 10/ (1 conf/machine/poodle-2.6.conf): poodle-2.6.conf: Add more modules, taken from .dev Jul 23 21:48:29 nite guys Jul 23 21:48:36 n8 zecke Jul 23 22:11:32 * CoreDump|home rbuilds poodle from scratch... Jul 23 23:05:08 Where's the code that drives the inherit something in the bb files? Jul 23 23:09:03 For instance one package has autotools binconfig lib_package, where is it defined what they do? Jul 23 23:16:07 Maybe they don't do anything... Jul 24 00:13:42 night **** ENDING LOGGING AT Mon Jul 24 02:59:57 2006